VARIEGATED, ORANGE AND BUTLER GLACIERS 121 change in the locus of emergence of drainage, the water was withdrawn from the western half of the fan allowing the development of an alder thicket. The maturity of the alders, indicates that the change in conditions occurred not less than a quarter of a century earlier. The cause for this change in drainage was inferred to be recession of the glacier, permitting a new point of outflow for the main drainage of the southern portion of the Variegated Glacier, and the natural inference was that this change was associated with the passage of the drainage through the granite gorge into the interior flat and thence out through the ice gorge upon the western margin of the alluvial fan. The accuracy of this interpretation was verified in 1906, for then, by the closing of the granite gorge, the southern marginal drainage was forced to flow along the southern margin of the glacier and emerge upon the eastern side of the alluvial fan. This new locus of emergence turned the distributaries of the glacial torrent down over the eastern half of the alluvial fan, as it had doubtless done in earlier time, and the alder thicket was being inundated and destroyed. In one of our expeditions we were obliged to cross this part of the alluvial fan, fording dozens of muddy torrents, and we witnessed a stage in the destruction of the alder thicket from near at hand. The bushes were being submerged in ice-cold water and in the rapidly accumulating deposits of coarse gravel which the glacial torrent was unable to move further. Many of the bushes were uprooted and were being whirled off toward the sea; others were dying from the long-continued submergence beneath the ice waters; and others were being killed by the deep burial of their roots beneath, the accumulating gravels. All the bushes were in full leaf, proving that the change had occurred during the growing season of 1906. The Valley Portion in 1909. On our visit in the summer of 1909, we found that the advance of the Variegated Glacier had entirely ceased, and that here, too, ablation had healed the surface so that travel over it was again possible (PL LV). The surface was still very rough, and there were many crevasses, but it seemed probable that one could go up the glacier as far as we went in 1905, though with much greater difficulty.1 We did not attempt this, though we went out to the center of the glacier, where it emerges from the mountains, and obtained a clear idea of the condition both in its upper and lower portions. One photograph, taken from a point not far away from the place where the red moraine stood in 1905, shows the .difference in condition very clearly. In 1905 there was a smooth, dirt-stained glacier surface, terminating abruptly in a wall of red moraine which rose on the down-stream side (PL LITE). All below t-Tiia was a rough, moraine-covered ice surface with the concentric bands of moraine already described; above, all was smooth ice with the exception of an occasional crevasse and a few small moraine ridges. In 1909, on the other hand, the clear ice area extended down much farther (PL LIV), the banded moraines were gone, and the ice surface consisted of a confused series of elevations and depressions with many shallow crevasses. We found it necessary to select our route with care where in 1905 we could go in almost any direction with equal ease. But, different as the surface was from the condition in 1905, it was even more strikingly different from its condition in 1906. There are two other notable contrasts between the 1905 and 1909 condition of this glacier. In the former year there were marginal valleys where the glacier emerged from its mountain valley; but in 1909 these were gone and the ice was in contact with the 1 In 1011 a Boundary Survey party went up Variegated Glacier to its very head and found no difficulty whatever in traveling over the surface which was impassably crevassed in 1906.